A multivariate statistical approach in correlating the acoustic properties with petrophysics and mineralogy on sandstones

Abstract

SUMMARYWe examined the effects of the petrophysical and mineralogical variables on the acoustic properties of Upper Assam sandstones. All parameters were evaluated in the laboratory using recognized standard laboratory methods. Compressional wave velocity was measured in the laboratory using piezoelectric transducers of 54 kHz by using the ultrasonic-through transmission technique. Single parameter correlations among bulk density, porosity, permeability and mineralogy with compressional wave velocity showed that the compressional wave velocity correlated inversely with porosity, permeability and feldspar content and directly with bulk density and quartz content. For instance, a plug with a higher amount of feldspar content showed a corresponding decrease in compressional wave velocity. Similarly, higher quartz content showed a higher compressional wave velocity. Nevertheless, to depict a clear correlation among different parameters, a multiparameter analysis was performed. It was observed that the coefficient of determination improved from 0.596 to 0.899 when compressional wave velocity was modelled in terms of bulk density, porosity, quartz and feldspar content collectively, rather than when compressional wave velocity was modelled as a function of porosity alone. Hence, our study suggests that multiple parameters exhibit a coherent influence on compressional wave velocity. Therefore, a multivariate statistical approach involving petrophysical and mineralogical influences would be more realistic. Furthermore, major mineral fractions of sandstone rock type, that is quartz and feldspar fractions were varied for a range of porosity to determine the variations in compressional wave velocity for different saturation conditions. Capturing the fluctuations in compressional wave velocity within the geological constraints will aid in possible reservoir characterization away from the well-control or beyond the available data range. The integrated approach can result in more accurate and precise rock physics models that may help to infer key reservoir parameters. The developed models can be used in quantitative seismic interpretation for reservoir characterization and to identify prospective reservoirs.